American Journal of Surgical Pathology 2017 June [Epub ahead of print] [Link]
Berg KB, Churg A
The separation of sarcomatoid and desmoplastic malignant mesotheliomas from sarcomatoid carcinomas of the lung metastatic to the pleura may be difficult, since both types of tumor can be morphologically similar and are frequently positive only for pan-keratin. GATA binding protein 3 (GATA3) is most commonly used as an immunohistochemical marker of breast and urothelial carcinoma, but is also known to stain other types of tumors including some mesotheliomas. In this study we asked whether GATA3 stains could be used to distinguish sarcomatoid/desmoplastic malignant mesotheliomas (N=19) from sarcomatoid carcinomas of the lung (N=13). Tumor staining was scored for diffuseness and intensity, with a maximum possible score of 6. All 19 sarcomatoid/desmoplastic malignant mesotheliomas examined showed strong diffuse staining for GATA3 (no case scored <3, mean score±SD for all 19 cases 5.4±0.9), whereas only 2 of 13 sarcomatoid carcinomas of the lung stained positively for GATA3 and the staining was weak and patchy (score 2 for each case, mean±SD for all 13 cases 0.4±0.8). There was no correlation between the intensity and diffuseness of GATA-3 staining and staining for traditional mesothelioma markers. Overall, any positive staining for GATA3 was 100% sensitive and 85% specific for sarcomatoid/desmoplastic mesothelioma. We conclude that strong diffuse staining for GATA3 favors a diagnosis of sarcomatoid/desmoplastic malignant mesothelioma over metastatic sarcomatoid carcinoma of the lung; conversely, complete absence of GATA-3 staining is evidence against a diagnosis of sarcomatoid/desmoplastic malignant mesothelioma.